CVE-2026-24233: CWE-502 Deserialization of Untrusted Data in NVIDIA TensorRT-LLM
NVIDIA TensorRT-LLM for Linux has a deserialization vulnerability in its restricted unpickler used for model weight deserialization. This flaw allows a local, unauthenticated attacker to deserialize untrusted data, potentially leading to code execution, privilege escalation, data tampering, and information disclosure. The vulnerability is identified as CVE-2026-24233 with a high severity rating and a CVSS score of 8.4. No official patch or remediation guidance has been provided yet by the vendor.
AI Analysis
Technical Summary
CVE-2026-24233 is a deserialization of untrusted data vulnerability (CWE-502) in NVIDIA TensorRT-LLM for Linux. The issue exists in the restricted unpickler component responsible for deserializing model weights. An attacker with local access and no authentication can exploit this flaw to execute arbitrary code, escalate privileges, tamper with data, or disclose sensitive information. The vulnerability has a CVSS 3.1 base score of 8.4, reflecting high impact on confidentiality, integrity, and availability. As of the published date, no patch or official remediation level has been disclosed by NVIDIA.
Potential Impact
Successful exploitation can result in arbitrary code execution, escalation of privileges, unauthorized data modification, and information disclosure on affected systems running NVIDIA TensorRT-LLM. The attacker requires local access but no authentication, increasing the risk in multi-user or shared environments.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is released, limit local access to trusted users only and monitor for suspicious activity related to TensorRT-LLM usage. Avoid loading untrusted model weights to reduce exposure.
CVE-2026-24233: CWE-502 Deserialization of Untrusted Data in NVIDIA TensorRT-LLM
Description
NVIDIA TensorRT-LLM for Linux has a deserialization vulnerability in its restricted unpickler used for model weight deserialization. This flaw allows a local, unauthenticated attacker to deserialize untrusted data, potentially leading to code execution, privilege escalation, data tampering, and information disclosure. The vulnerability is identified as CVE-2026-24233 with a high severity rating and a CVSS score of 8.4. No official patch or remediation guidance has been provided yet by the vendor.
CVSS v3.1
Score 8.4high
Affected software
pkg:github/nvidia/TensorRT-LLMRun on your own infrastructure? Check whether these packages are installed with threat-finder — our free open-source scanner.
Weaknesses
AI-Powered Analysis
Machine-generated threat intelligence
Technical Analysis
CVE-2026-24233 is a deserialization of untrusted data vulnerability (CWE-502) in NVIDIA TensorRT-LLM for Linux. The issue exists in the restricted unpickler component responsible for deserializing model weights. An attacker with local access and no authentication can exploit this flaw to execute arbitrary code, escalate privileges, tamper with data, or disclose sensitive information. The vulnerability has a CVSS 3.1 base score of 8.4, reflecting high impact on confidentiality, integrity, and availability. As of the published date, no patch or official remediation level has been disclosed by NVIDIA.
Potential Impact
Successful exploitation can result in arbitrary code execution, escalation of privileges, unauthorized data modification, and information disclosure on affected systems running NVIDIA TensorRT-LLM. The attacker requires local access but no authentication, increasing the risk in multi-user or shared environments.
Mitigation Recommendations
Patch status is not yet confirmed — check the vendor advisory for current remediation guidance. Until an official fix is released, limit local access to trusted users only and monitor for suspicious activity related to TensorRT-LLM usage. Avoid loading untrusted model weights to reduce exposure.
Technical Details
- Data Version
- 5.2
- Assigner Short Name
- nvidia
- Date Reserved
- 2026-01-21T19:09:37.972Z
- Cvss Version
- 3.1
- State
- PUBLISHED
- Remediation Level
- null
Threat ID: 6a569d1c68715ace432809cf
Added to database: 07/14/2026, 20:33:32 UTC
Last enriched: 07/21/2026, 22:34:59 UTC
Last updated: 08/28/2026, 22:52:09 UTC
Views: 92
Community Reviews
0 reviewsCrowdsource mitigation strategies, share intel context, and vote on the most helpful responses. Sign in to add your voice and help keep defenders ahead.
Want to contribute mitigation steps or threat intel context? Sign in or create an account to join the community discussion.
Actions
Updates to AI analysis require Pro Console access. Upgrade inside Console → Billing.
Need more coverage?
Upgrade to Pro Console for AI refresh and higher limits.
For incident response and remediation, OffSeq services can help resolve threats faster.
Latest Threats
Check if your credentials are on the dark web
Instant breach scanning across billions of leaked records. Free tier available.